.png)
Ask three agencies to quote a Shopify rebuild for a store carrying 4,000 SKUs across wholesale and retail pricing, and you'll get three very different numbers. The lowest one is the one that should worry you most.
A low commerce quote often prices a different job than the one you asked for: a themed template with your logo, colours and a handful of custom sections layered on top. That can look like a full commerce build in a pitch deck. It behaves like one for a few months, until a variant matrix outgrows the theme's assumptions or a wholesale customer needs pricing the theme was never built to show.
If you've read our guide to ecommerce in NZ, treat this as the companion piece: the specific questions to ask once you've decided a rebuild or a new build is the right move. General website vetting (do they answer emails, do they show up on time, do you like their portfolio) still matters. It just isn't enough for commerce.
Commerce sites carry inventory, tax rules, freight logic and payment providers that a brochure site never has to touch. The agency you choose needs to have actually handled that weight before, not just designed around it.
The theme-with-a-logo problem
Shopify makes it easy to launch something that looks finished fast. Pick a premium theme, swap the fonts and colours, load your products and the store can look like a six-figure build within a fortnight. For a small catalogue with simple pricing, that might be the right call.
For an established business with real inventory complexity, it's a different story. Themes are built around assumptions: one price per variant, one warehouse, one currency, a manageable number of products. Push past those assumptions and you're either paying an app to patch the gap or asking the theme to do something it was never designed for.
Ask any agency you're considering to walk you through a store they've built that had genuine complexity behind it, not just a nice colour palette in front of it. Ask what the catalogue size was, how pricing varied by customer type and what happened when the business needed something the theme didn't ship with. If the answer is vague, that's the answer.
Inventory complexity: the question generic vetting skips
Most agency comparisons stop at design sensibility and project management style. For commerce, the sharper question is whether they've built for real stock and pricing structures: multiple warehouses, wholesale tiers alongside retail pricing, bundled products, pre-orders, or stock that needs to sync against a separate inventory system in near real time.
Clean Collective is a useful benchmark here. Its rebuild had to handle stockist mapping and a 3PL fulfilment integration across the New Zealand and Australian markets at once, and the site now converts 103% better than before, with sales up 28%. That result came from the commerce architecture underneath the design, not from a sharper hero image.
Our Shopify website design work and, for larger catalogues and multi-entity setups, Shopify Plus builds are both scoped around this kind of complexity from the first strategy conversation, not patched in once the theme runs out of road.
App strategy and the cost of app sprawl
Every Shopify store accumulates apps. A subscription tool here, a reviews widget there, a bundling app because the last one didn't quite fit. Each one seems cheap and reasonable on its own. Three years in, a store can be running twenty-plus apps, several of which nobody remembers installing, all billing every month regardless and all adding weight to page speed.
Ask a prospective agency how they think about app strategy up front rather than after the fact. Do they map which functions actually need a dedicated app, which can be handled natively in Shopify and which are duplicating something another app already does? An agency that can't answer that question in a sales call is unlikely to raise it once the invoice is paid. We've written a full breakdown of how to clean up app sprawl without breaking anything if you want the detail on how that audit actually runs.
Build versus buy: how they make the call
Nearly every commerce feature can be solved two ways: buy an app that does it, or build the logic into the site. Neither answer is automatically right. An app is faster to switch on and comes with vendor support, but it adds a recurring cost and another dependency to manage. A built feature costs more upfront and sits entirely inside your own site, with no third party able to change pricing or shut the feature down.
A good agency treats this as a case-by-case decision, weighing transaction volume, how core the feature is to your margin and how long you're likely to need it, against the ongoing app fee. A weaker one defaults to whichever app they're used to installing, because it's faster for them, not because it's cheaper for you over three years. Ask directly: for a feature like subscriptions, loyalty points or gift cards, how would they decide whether to buy it off the shelf or build it in?
Migration experience with revenue still live
Platform migrations are where inexperience gets expensive fastest. Moving a store from Magento, WooCommerce or an older Shopify setup to a new platform means the old site has to keep taking orders right up until the new one goes live, with product data, order history, customer accounts and SEO value all carried across intact.
Our Advintage migration moved more than 5,000 SKUs from Magento to Shopify, including complex discount structures, wholesale pricing and an ERP connection, with zero downtime across the switch. That's the standard to ask for: not "have you done a migration", but "have you done one where the business kept selling the entire time, and what would you have done differently".
Ask, too, what happens to your current URLs and product data. A migration that loses ten years of accumulated search rankings because redirects weren't mapped properly is a hidden cost that shows up months later, in traffic reports rather than in the invoice.
Freight, tax and payment: the NZ specifics
New Zealand commerce has its own mechanics, and an agency that's only built for other markets can miss them. GST needs to be handled correctly at checkout, including for businesses selling into Australia where the tax treatment differs. Freight calculation runs through NZ Post, Starshipit or Go Sweet Spot, each with its own rate logic and rural delivery rules that a generic freight app doesn't always price correctly out of the box.
Payment expectations matter too. NZ shoppers expect Afterpay or Laybuy alongside card payment, and providers like Windcave or POLi still see real volume here, particularly for higher-value B2B orders. If you sell into Australia as well, dual-currency pricing and checkout logic need to be right from day one, not retrofitted once the first AUD order causes a support ticket.
None of this is exotic. It's ordinary NZ commerce reality, and it's exactly the kind of detail that separates an agency that's built here from one that's building here for the first time.
Integrations: who actually connects what
Most established businesses already run an accounting system, an ERP or an inventory platform, and the new site needs to talk to it. A good agency maps that integration properly during strategy, working out exactly how your existing systems, whether that's Xero, an ERP like Infusion or Cin7, or a 3PL, will function within the new site.
Skyrocket scopes, manages and facilitates these connections rather than writing the underlying integration code ourselves. That technical build sits with the platform vendor or the relevant app, and a straight answer about who does what protects you from an agency claiming ownership of work another partner is actually doing. Our piece on complex Shopify integrations for ERP, inventory and accounting covers how that mapping and management process runs in practice, and what to ask before you commit to a scope.
Common questions
How do I know if an agency has really built complex commerce sites?
Ask for specifics: SKU counts, variant depth, whether pricing varied by customer type and what integrations were involved. A specific answer with numbers is a good sign. A general answer about their design process often means the complexity wasn't there.
Is a themed Shopify site ever the right choice?
For a small catalogue with simple, uniform pricing and no wholesale or multi-warehouse requirements, a well-chosen theme with sensible customisation can work well. The risk is being sold that approach for a business that has already outgrown it.
What should a migration quote actually include?
It should cover full product and customer data migration, order history handling, redirect mapping for existing URLs and a plan for keeping the old site selling until the new one is confirmed live. If a quote doesn't mention redirects or downtime at all, ask why.
How many apps is too many for a Shopify store?
There's no fixed number. The better question is whether each app earns its monthly fee and its share of page load time. A proper app audit, covered in our app sprawl guide, will often surface a handful doing nothing.
Do I need Shopify Plus, or is standard Shopify enough?
It depends on order volume, the number of sales channels and how much custom checkout logic you need. Businesses with multiple entities, complex B2B pricing or very high transaction volumes tend to outgrow standard Shopify's limits. Our Shopify Plus page sets out where that line tends to sit.
Who handles integrations with our accounting or ERP system?
The agency should map the integration strategy and manage the connection during the build. The underlying technical work sits with the platform vendor or the relevant app, with the agency scoping and coordinating it rather than writing that code itself.
If your current site is carrying more complexity than the last agency accounted for, or you're about to sign a quote that looks suspiciously simple, talk to us. Our ecommerce website build team can walk through your specific inventory, integration and platform questions before you commit to anyone. Get in touch and we'll tell you plainly whether what you're being sold matches what your business actually needs.
